The paper presents a method for determining (estimating) the state variables of an electronically commutated motor (ECM). For the estimation of state variables and motor parameters, the extended Kalman filter method was applied, using the signals of only those electric motor variables that are available for measurement (stator line voltages and stator currents). It was assumed that the measurement signals of voltage and stator current are not filtered in real operation, which is otherwise common when applying similar methods. The mean value of the voltage within the sampling interval is determined by combining imaginary measurements and calculations, thanks to the application of a predictive current regulator. The reliability of the estimate was checked by computer simulation, for which two mathematical models were created. Using one model, the existing drive is simulated (without variable estimation) and the values of the EKM state variables are determined. The second model, which determines the estimated values of engine variables and parameters, uses an extended Kalman filter and requires the definition of a discrete engine model. The simulation results (in addition to the expected accuracy of the estimation of directly measured currents of the EKM) show a good accuracy of the estimation of the speed of rotation and the position of the rotor in the dynamic and static modes of operation of the EKM.